The Base Needed 1.9 Million Cubic Metres Out. The Hill Lost 41,000.
Two survey firms, working blind, computed seven decades of volume change from film and lidar this newspaper paid for. The hill has lost 41,000 cubic metres, all of it in one road cut on the north side.
A standing series in which this newspaper investigates a widely held claim and reports that it could not substantiate it.
ABERGAVENNY, Wales — Twelve levels under Mynydd Cadlan would require about 1.9 million cubic metres of rock to be cut out of it.
Broken rock swells. Hauled away, that becomes roughly 2.8 million cubic metres of muck, which is about 93,000 lorry movements on two lanes that carry weight limits — and 93,000 is the generous figure, calculated at 30 cubic metres a load. A tipper that can actually get up those lanes carries nearer twelve.
Between 1953 and 2024 the hill lost 41,000 cubic metres.
All of it is in one place, on the north side, in a shape that is unmistakably a road cut.
Three surfaces, bought and published
We did not ask whether there was a base. We asked how much rock had left the hill.
Three surfaces were purchased. The 1953 and 1971 aerial survey sorties were scanned at 12 microns from the original negatives, not from prints, which cost five times as much and was the single most important decision in the reporting. The third is an airborne lidar pass flown to order in October 2024.
Two survey firms then computed the volume change between them. Neither knew about the other. Neither was told the hill's name or the reason for the job until its report was delivered; both were told it was a landfill volumetric dispute.
Pellworth Geospatial returned a net loss of 40,100 cubic metres, plus or minus 3,400. Cadair Survey Partnership returned 42,600, plus or minus 2,900. They disagree for six pages about the 1971 film's interior orientation, and both reports run in full with this story.
The trucks were real
The most striking thing about this story is that the traffic people remember actually happened, in the volumes they remember, in the years they remember.
Between March 1968 and August 1971, a Monmouthshire county highways contract moved 38,400 cubic metres of shale out of the north cut and built it into an embankment eleven miles away. The tender, the measured works, the weighbridge tickets, three contractual claims and a dispute about Sunday working are in the county record office in nine box files. Our correspondent read them over eleven days and every haulage ticket is transcribed and published with this piece.
The contractor's daughter still has her father's day book. It records the dust, which was orange and got into the bread; two broken leaf springs in one week in June 1969; and what Idris Prydderch thought about being asked to run on a Sunday, which was a great deal.
Three years of genuine, memorable, unpleasant traffic. It matches the road cut in the lidar to within an amount two survey firms would argue about. It does not touch 93,000.
One room
There is an underground structure at the centre of the story, and it exists.
It is a covered service reservoir. Nine hundred cubic metres, built in 1954, with a concrete roof under about a metre of turf, an access hatch, a fixed ladder and a padlock. The undertaker's asset register gives its dimensions. The visit log begins in 1998.
Nine hundred cubic metres against 1.9 million is a ratio of about one to two thousand. The room is real. The eleven floors were added somewhere between the hatch and the telling.
What would defeat this test
We will state it plainly rather than leave it for a correspondent to find.
Our method measures the outside of a hill. If spoil was driven out sideways through an adit and tipped into the worked-out void of the disused Blaenserw drift, two valleys over, it would never cross the surface we surveyed and this entire article would be worthless.
We obtained Blaenserw's subsidence monitoring from 1961 to 2019. There is no settlement anomaly, no inflection in the record, and no change in the pumping regime.
That is not the same as proving nothing went in. The drift was levelled twice a year by one man, and a hundred thousand cubic metres arriving slowly is exactly the shape of thing that record would miss.
What this desk takes from it
Emlyn Cadwaladr-Nunn's reply is printed below at his requested length, and it takes something from us.
He never said twelve; a television programme did. He says the finding is not the volume but the four-line refusal that a ministry has sent to this valley for fifty-one years about a hatch containing drinking water.
We measured the hill because the hill could be measured. We have no method at all for the four lines.
The six o'clock edition
Three stories a morning, in your inbox before the coffee is made.
Sources & Method
We did not ask whether there was a base. We asked how much rock had left the hill, which is a smaller question with an answer. Three surfaces were bought — 1953 film, 1971 film and a 2024 lidar sortie flown to order — and two survey firms computed the change between them without knowing why, without knowing about each other, and without being told the hill's name until the reports were in.
Who we spoke to
- Pellworth Geospatial, Photogrammetric reconstruction, 1953 and 1971 survey film. Commissioned by this newspaper; told the job was a landfill volumetric dispute and given no other context until after delivery February–May 2026 Reports a net loss of 40,100 m³ ±3,400. Its report is published entire, including the eight control points it could not recover.
- Cadair Survey Partnership, Second reconstruction, blind to the first. Given the identical scans and the same cover description, with no sight of the first report and no contact with its author March–June 2026 Reports 42,600 m³ ±2,900. Disagrees with Pellworth about the 1971 film's interior orientation and says so for six pages.
- 1953 and 1971 aerial survey sorties, Original negatives, national air photograph collection. Scanned to order at 12 microns from the negatives rather than from prints; scans paid for and republished by us January 2026 Prints would have cost a fifth as much and would have carried the print's own distortions into every number in this article.
- Monmouthshire county highways contract file, 1967–1972, Nine box files: tender, measured works, weighbridge tickets, three claims and a dispute about Sunday working. Read at the county record office over eleven days; every haulage ticket transcribed and published as a spreadsheet April 2026 38,400 cubic metres of shale, cut to embankment, eleven miles. The trucks people remember.
- Gwenllian Prydderch-Amis, Daughter of Idris Prydderch, haulage contractor. Interviewed at her home twice; lent us her father's day book for nine days and read the full draft before publication May and July 2026 The day book records the dust, the broken springs, and what he thought of working Sundays. It is the only voice in this story that was there.
What we could not confirm
- Whether anything was tipped into the disused Blaenserw drift two valleys over. That is the one route by which spoil leaves a hill without touching a public road, and it would defeat this entire test. We obtained the drift's subsidence monitoring, 1961 to 2019, and found nothing anomalous — no settlement, no inflection, no change in the pumping regime. That is not the same as proving nothing went in. Monitoring that sparse would not see a hundred thousand cubic metres arriving slowly.
- The 1953 surface's control. Photogrammetry from historic film needs ground control, and we have 61 points from the survey's own trig list. Eight of them no longer exist. Both firms handled the gaps differently and their two answers differ by 2,500 cubic metres, which is the honest size of our uncertainty.
- Whether the reservoir has ever been entered by anyone other than the undertaker's own staff. The padlock is a water company padlock and the visit log begins in 1998. Before that there is no log, and the men who would have kept it are dead.
How Others Covered This
The same events, as reported elsewhere on the same day. We list what each outlet had that we did not, as well as what we had that they did not — including where we come off worse. Why we print this.
-
The Hollow PostTHE HILL THEY WILL NOT LET YOU SURVEY
Built the piece on a ministry's four-line refusal to confirm or deny, and treated the refusal as the finding.
Had that we did not
The refusal itself, in full, which is genuinely the most interesting document anybody has produced about this hill in fifty years.
Left out
That we surveyed it, from the air, twice, at our own expense, published the point cloud, and that nobody attempted to stop us.
-
ApertureThe Shape Of A Rumour In A Valley
An essay on what a post-industrial valley does with a hill it can no longer work. Better written than this piece and not attempting the arithmetic.
Had that we did not
The observation that the story arrived in the valley from outside it, on television, and was then adopted.
Left out
Idris Prydderch's day book, which is the only contemporaneous account of the traffic anybody remembers and sits eleven miles away in a kitchen drawer.
-
They Buriedthis newspaperThe Base Needed 1.9 Million Cubic Metres Out. The Hill Lost 41,000.
Bought three independent surfaces of the same hill seventy-one years apart and had the volume change computed twice, blind.
Had that we did not
Both firms' reports entire, the point cloud, the disagreement between them, and the failure mode that would defeat the whole test.
Left out
We spent £41,000 answering a question no official body would answer for fifty years, and then wrote a piece about how the answer is dull. Mr Cadwaladr-Nunn says the refusal is the story and this desk has not shown that he is wrong. — V. Ashcombe-Doyle, standards editor
Right of Reply
They Buried contacted Emlyn Cadwaladr-Nunn, who has been describing the hill in print and on air since 1979 on 9 July 2026 in person at his home, again on 24 July with the full draft, both survey reports and the point cloud on a memory stick, and four weeks to answer. Replied by letter on 14 August 2026, running to eleven pages, and asked that it not be cut. Printed unedited.
I have never said twelve. I want that first because you have put it in your headline and it is not mine.
What I said, in 1979, in a pamphlet of which four hundred were printed, was that the hill had been worked between 1968 and 1971 by men who were not from the valley, that the works were larger than a road cut required, and that nobody in authority would say what they were. Twelve levels came from a television programme in 1994 that did not interview me and did not need to. I have spent thirty years being asked to defend a number I did not produce, and now a newspaper has spent forty-one thousand pounds disproving it and has put my name under it.
That is my first complaint and it is a small one. Here is the larger one.
Your method measures the outside of a hill. It is very good at that and I have no quarrel with your two firms, whose reports I have read twice and one of which I think is the better piece of work. But you have assumed that rock which is removed must go up and out over the top, and that is not how anyone has ever driven an adit in this country. You drive in from the side, you take the spoil out level, and if there is a worked void within a few miles you put it there, because tipping is a cost and a hole is free. You know this. You have written it into your own list of failure modes and then declined to let it disturb your conclusion.
You checked the subsidence monitoring at Blaenserw and found nothing. I would not expect you to. That drift was monitored twice a year by a man on a motorbike with a level staff, and I knew him.
Now the thing I actually want said.
For fifty-one years I and others have written to the ministry, to the county, to two water undertakers and to four members of Parliament, and the answer has always been the same four lines: it is not the department's practice to comment on the location or function of sites of this kind. Not 'there is nothing there'. Not 'it is a reservoir'. The same four lines, for fifty-one years, about a padlocked hatch in a Welsh hillside that your own article says holds nine hundred cubic metres of drinking water.
If it is a reservoir, why has no one been permitted to say so?
You will answer that the four lines are a standing form, applied without thought, and you are probably right, and that is precisely the point I have failed to make for fifty years. A state that will not distinguish between a bunker and a water tank has told the valley that both are the same kind of secret. It taught us to read a padlock as a message. Then a television programme arrived with a number, and the valley had already been trained to believe it.
I was wrong about the size. I was not wrong about the silence, and your paper has now been silent about the silence for nine hundred words.
One more thing, and then I have finished. You have printed Idris Prydderch's day book and you have printed it well. He was a good man and a rough employer and my uncle drove for him for two years. But I would ask you to notice what you have done with him: you have used a working man's account of three years of dust to close a question, and you have used it because it agrees with you. Had his day book said something else, you would have called it recollection.
Published unedited under our right-of-reply guarantee.
How was this story?
We publish the result, whatever it is. Reader verdicts appear on the front page and in our newsroom metrics.
Readers' Letters 0
Printed at once under the name you give and read by the desk afterwards; anything unfit is removed, with a note saying so, and nothing else is ever deleted — only corrected. Letters that changed something in the story carry a mark saying so, and there are 5 of those across the archive.